Benchmarking and systems performance - hosted by Elisabeth Stahl

Elisabeth Stahl

Blog Authors:  ELISABETH STAHL   is Chief Technical Strategist and Executive IT Specialist, IBM Systems and Technology Group, and has been working in systems performance for over 25 years.

  • Add a Comment
  • Edit
  • More Actions v
  • Quarantine this Entry

Comments (3)

1 Kleber commented Permalink

Thanks Elisabeth<div>&nbsp;</div> I wrote about it in my blog. I didn't know Solid-state Storage technology web site (http://www-03.ibm.com/systems/storage/solutions/ssd/).

2 BobTheDataBaseBoy commented Permalink

I have read a bit of your and Barry Whyte's blog on SSD. In none of these postings could I find any discussion of the factor which makes SSD truly valuable (beyond raw I/O): since random R/W is independent of data location, unlike HDD, database joins are marginally costless. This fact leads to the realization that SSD mated to multi-core/processor machines are the "perfect" platform to implement BCNF datastores ("xml??? we don't need no stinking xml"). Such datastores can be an order of magnitude or two smaller than their flatfile equivalents, and will run faster with the concomitantly lesser I/O load. You might sell fewer drives, but hey, it's better for the customer.<div>&nbsp;</div> Any notion who might be working this area? Either in storage or DB2? If no one is, I'm available.

3 holsonw commented Permalink

Bob,<br /> Thanks for your comment. I do agree that the random access patterns of complex queries are perfect for SSDs and multi-processor machines. We haven't done exhaustive testing in the Benchmark Center but some of the workloads we have run have proven this out as well. <br /> Helen Olson-Williams

Add a Comment Add a Comment